Shows the layout of samples, controls (e.g., negative controls (NC), Calibrator controls (CAL) and special wells if any (e.g., AQ sample controls (AQSC)) on the 96 well assay plate.
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| A | X252L | X268L | X254L | X270L | X315L | X331L | X317L | X333L | X284L | X300L | X286L | X302L |
| B | X253L | X269L | X255L | X271L | X316L | X332L | X318L | X334L | X285L | X301L | X287L | X303L |
| C | X256L | X272L | X258L | X274L | X319L | X335L | X321L | X336L | X288L | X304L | X290L | X306L |
| D | X257L | X273L | X259L | X275L | X320L | NC 3 | X322L | NC 1 | X289L | X305L | X291L | X307L |
| E | X260L | X276L | X262L | X278L | X323L | NC 4 | X325L | NC 2 | X292L | X308L | X294L | X310L |
| F | X261L | X277L | X263L | X279L | X324L | AQSC 2 | X326L | CAL 2 | X293L | X309L | X295L | X31AL |
| G | X264L | X280L | X266L | X282L | X327L | AQSC 3 | X329L | CAL 3 | X296L | X311L | X298L | X313L |
| H | X265L | X281L | X267L | X283L | X328L | AQSC 1 | X330L | CAL 1 | X297L | X312L | X299L | X314L |
Summary of the number and classification of reads for the assay run
Each sample of the assay is spiked with the same concentration of at least one internal control (IC). This is used both as part of a well normalization procedure and as a method for assessing the uniformity of the assay run. SD: standard deviation. CV: coefficient of variation
Internal Control (IC) summary - mCherry
CAL samples are pooled plasma controls used both to normalize samples between experiments and to assess performance (e.g., precision) of the assay.
Note: %CV in this table does not exclude values below LOD. See Coefficient of variation section for %CV (before and after normalization) that excludes values below LOD.
IPC summary
NC wells are reactions where no sample input is provided (i.e., buffer only). These are used to assess Limit of Detection (LOD) for each target assay and to assess run quality (e.g., background levels).
NC summary
The following values for NC wells and targets were identified as outliers and excluded from LOD calculations.
AQSC samples are pooled plasma controls from an independent pooled plasma source that is different from CALs. AQSCs are used for calculating intra- and inter-plate coefficient of variation (CV). They are used to obtain an unbiased measure of inter-plate CV when CAL normalization is used.
Note: %CV in this table does not exclude values below LOD. See Coefficient of variation section for %CV (before and after normalization) that excludes values below LOD.
AQSC summary
Heatmaps show the percent relative to plate median for log2(total counts) and the specified internal control(s).
Plate-specific QC criteria:
QC criteria are assessed either on unnormalized (raw) or CAL-normalized (CAL) data.
Number of warnings: 9
Combined Plate QC Summary
Sample-specific QC criteria:
Number (%) of samples that have warnings for each QC criterion (excludes CAL, AQSC & NC samples)
Wells with warnings: Detectability
Wells with warnings: IC Reads
Wells with warnings: Reads
Wells with warnings: IC Median
Target-specific QC criteria:
Number of Targets with warnings: 227
Number (%) of targets that have warnings for each QC criterion
Accuracy
CV
For top 10 targets by % reads on each plate, the table shows target name, % of grand total plate reads, and total target reads. Note that all targets are considered including any excluded or IC targets.
Top 10 targets by % reads
For top 10 samples by % reads on each plate, the table shows sample name, % of grand total plate reads, and total sample reads. Note that all samples are considered including any excluded or control samples.
Top 10 samples by % reads
Target detectability is the percentage of samples that are above the limit of detection for that target. A target is considered “detectable” if it is above limit of detection in greater than 50% of samples. “Overall” detectability is the overall percentage of samples across all plates that are above LOD for a given target. Target detectability reported here excludes CAL, AQSC, & NC samples.
Plate_01
Plate_01
Plot shows target IC-CAL-normalized count distributions relative to LOD. For each target, log2(LOD) was subtracted from log2(count) for IC-normalized data. Reverse-curve quantified targets are denoted by a “*” and are plotted as forward-curve targets for compatibility
Target quantifiability is the percentage of samples that are within the dynamic range, that is, above the lower limit of quantification (LLOQ) and below the upper limit of quantification (ULOQ) for that target. A target is considered “quantifiable” if at least 50% of samples fall within the dynamic range. “Overall” quantifiability is the overall percentage of samples across all plates that are within the dynamic range for a given target. Target quantifiability reported here excludes CAL, AQSC, & NC samples.
overall
PLASMA
CSF
SERUM
OTHER
%CV was calculated for each target (excluding internal controls) for the AQSCs and CALs. Only values above LOD were used to calculate %CV. Some targets may have missing CV due to too many values below LOD.
AQSC intra-plate %CV -- unnormalized
CAL intra-plate %CV -- unnormalized
AQSC intra-plate %CV -- IC-normalized
CAL intra-plate %CV -- IC-normalized
CVs reported here were calculated on the aM units AQ data. Based on 228 AQ targets.
AQSC intra-plate %CV -- AQ
CAL intra-plate %CV -- AQ
Targets with any intra-plate CV > 30% are highlighted in red.
Number of warnings: 6
AQSC intra-plate CV, IC-normalized
Plot shows the log2-transformed ratio (left) or the unlogged ratio (right) of the mean IC-normalized counts for AQSC vs CAL.
Plot shows the log2-transformed ratio of the mean IC-normalized counts for AQSC versus NC. Asterisk next to target name indicates the NC mean was zero for one or more plates, so AQSC / NC ratio could not be calculated for these plates.
Sample boxplots show distributions of the log2 counts for each sample, unnormalized and after IC + CAL normalization.
Plot shows the Pearson correlation between samples. Hierarchical clustering is done using complete linkage.
Hierarchical clustering of both samples and targets was done using a Euclidean distance metric and complete linkage. Data is log2 and IC-CAL normalized; targets are centered and scaled.